Apparatus and method for motion video encoding employing an adaptive quantizer

Abstract

An apparatus and method for encoding of a video picture is disclosed. The video picture has a plurality of pictures. The pictures each have a plurality of macroblocks. The macroblocks each have a plurality of sub-blocks. The apparatus comprising a first module configured to generate a transform coefficient C ij for each of the sub-blocks of the macroblock. The apparatus further comprises a second module configured to variably quantize the transform coefficient by a scaling factor Q p based on the complexity of the picture and any rate control requirements.

Having thus described our invention, what we claim as new and desire to secure by Letters Patent is: 
     
       1. A video encoding apparatus configured to convert groups of blocks of digital video signals into compressible groups of blocks of digital video signals, comprising: (a) a transform coefficient device, having an input configured to receive a digital video signal representing a block of a video image, configured to generate a first signal representing a set of transform coefficients C ij  in response to said received digital video signal;   (b) an adaptive quantization device, having an input coupled to said transform coefficient device, configured to receive said first signal representing a set of transform coefficients C ij , and to provide a second signal representing a quantized transform coefficient Cq ij  in response to said first signal representing said set of transform coefficients C ij  ;   (c) a variable length coder, having an input coupled to said adaptive quantization device, configured to receive said second signal representing said quantized transform coefficient Cq ij  and to provide a variable-rate signal having a variable bit rate and representing a compressed digital video signal in response to said second signal representing said quantized transform coefficient Cq ij  ; and   (d) a buffer, having an input coupled to said variable length coder, configured to equalize said variable bit rate of said variable-rate signal representing a compressed digital video signal and to provide a constant rate compressed digital video signal, and further configured to provide a third signal indicating buffer fullness.   
     
     
       2. The apparatus of claim 1, further comprising a bit allocator, coupled to said adaptive quantization device, configured to allocate a specific number of bits for each picture. 
     
     
       3. The apparatus of claim 1, further comprising a storage device, coupled to said buffer, configured to store said constant rate compressed digital video signal. 
     
     
       4. The apparatus of claim 1, wherein said second signal representing a quantized transform coefficient Cq ij  is further provided in response to said third signal indicating buffer fullness. 
     
     
       5. The apparatus of claim 4, wherein said second signal representing a quantized transform coefficient Cq ij  is further provided in response to complexity of the video image represented by said digital video signal. 
     
     
       6. The apparatus of claim 1, wherein said second signal representing a quantized transform coefficient Cq ij  is further provided in response to complexity of the video image represented by said digital video signal. 
     
     
       7. A video encoding apparatus configured to convert groups of blocks of digital video signals into compressible groups of blocks of digital video signals, comprising: (a) a transform coefficient device, having an input configured to receive a digital video signal representing a block of a video image, configured to generate a first signal representing a set of transform coefficients C ij  in response to said received digital video signal; and   (b) an adaptive quantization device, having an input coupled to said transform coefficient device, configured to receive said first signal representing a set of transform coefficients C ij , and to provide a second signal representing a quantized transform coefficient Cq ij  in response to said first signal representing said set of transform coefficients C ij , wherein said adaptive quantization device comprises: (1) a first multiplier configured to provide a third signal representing partially quantized video data CW ij  in response to said first signal representing a set of transform coefficients C ij  ;   (2) a q p  selection device, coupled to said first multiplier, operable to provide a fourth signal representing a quantization factor q p  in response to said third signal representing partially quantized video data C ij  ; and   (3) a second and third multiplier and an adder, configured to transform said third signal representing partially quantized video data C ij  into said second signal representing said quantized transform coefficient Cq ij , based on said fourth signal representing said quantization factor q p .     
     
     
       8. The apparatus of claim 7, further comprising: (c) a variable length coder, having an input coupled to said adaptive quantization device, configured to receive said second signal representing said quantized transform coefficient Cq ij  and to provide a variable-rate signal having a variable bit rate and representing a compressed digital video signal in response to said second signal representing said quantized transform coefficient Cq ij  ; and   (d) a buffer, having an input coupled to said variable length coder, configured to equalize said variable bit rate of said variable-rate signal representing a compressed digital video signal and to provide a constant rate compressed digital video signal, and further configured to provide a fifth signal indicating buffer fullness.   
     
     
       9. The apparatus of claim 7, wherein said fourth signal representing said quantization factor q p  is further provided in response to a signal representing fullness of a rate buffer. 
     
     
       10. The apparatus of claim 7, wherein said q p  selection device comprises: (a) a first max select module, having an input to receive said third signal representing partially quantized video data CW ij , configured to determine a maximum transform coefficient for each of four blocks of each group of blocks, and to generate a firth signal representing the same;   (b) a first min select module, having an input coupled to said first max select module, configured to determine a minimum transform coefficient for each four blocks of each group of blocks, and to generate a sixth signal representing the same;   (c) a normalizing multiplier, having an input coupled to said first min select module, configured to normalize said sixth signal representing said minimum transform coefficient for each of four blocks of each group of blocks, and further configured to provide a seventh signal representing the same;   (d) a q p  min selection module, having an input coupled to said first max select module, configured to select a value of q p  that is a lower bound for q p , and to provide an eighth signal representing the selected value of q p  min;   (e) a lower bound select module, having a first input coupled to said q p  min selection module and a second input coupled to said buffer, configured to provide a ninth signal representing a lower bound of q p  ;   (f) a second max select module, having a first input coupled to said normalizing multiplier and a second input coupled to said lower bound select module, configured to provide a tenth signal representing a value of q p  that is the greater of said seventh signal and said ninth signal; and   (g) a second min select module, having a first input coupled to said second max select module and a second input coupled to said buffer, configured to select a value for q p  that is not greater than rate control requirements allow and further configured to provide an eleventh signal representing this selected value of q p .